2.3 KiB

Raw Blame History

Log Chat Format

Purpose

log-chat.md tracks agent activity across iterations. Critical for:

@editor seeing @seo changes (understanding SEO edits vs errors)
@writer seeing @editor feedback (knowing what to fix)
@seo seeing previous optimization attempts
Any agent understanding context of prior work

When to Use

Before work: Read log-chat.md to understand previous iterations.

After work: Append your entry with what was done.

Entry Format

## [YYYY-MM-DD] @agent-name

**Action:** Brief description of what was done

**Changes:**
- file.md: what changed
- another-file.md: what changed

**Notes:** Important decisions, reasoning, observations

**For next agent:** Specific instructions or warnings

---

Examples

@editor entry

## 2025-01-07 @editor

**Action:** Review and feedback on draft

**Changes:**
- text.md: no changes yet, feedback only

**Notes:** 
- Introduction too long, suggest cutting first paragraph
- Code example in section 3 needs output screenshot
- Tone inconsistent in conclusion

**For next agent:** @writer please address feedback above

---

@seo entry

## 2025-01-07 @seo

**Action:** SEO optimization pass

**Changes:**
- text.md: added keywords to H2 headings, optimized first 100 words
- seo-metadata.md: created with title, description, keywords

**Notes:**
- Primary keyword "claude sandbox" added 3 times naturally
- Rewrote opening paragraph for SERP snippet optimization
- All images now have keyword-rich alt text

**For next agent:** Text changes are SEO-driven, not style errors

---

@writer entry (after iteration)

## 2025-01-08 @writer

**Action:** Applied editor feedback

**Changes:**
- text.md: shortened intro, fixed tone in conclusion

**Notes:**
- Kept SEO optimizations from @seo
- Added transition paragraph between sections 2-3

**For next agent:** Ready for final review

---

Key Principles

Be specific — "fixed intro" is useless, "removed first paragraph, added hook question" is helpful
Explain reasoning — especially for @seo changes that might look like errors
Direct next agent — say who should act and what they should focus on
Keep it brief — this is a log, not documentation

Updated: 2025-01-07

2.3 KiB Raw Blame History